Jan Vlk

Jan Vlk

Autonomous Systems Researcher

PhD student in Cybernetics & Robotics at CTU in Prague
Building perception systems for autonomous robots

6+ Years in robotics
4 Competition podiums
12 Projects

News & Updates

Sep 2026 Award

1st Place at Robotour 2026 — Our team won the main category of the international outdoor robot competition in Prague’s Stromovka, where I served as the lead integrator of the robot’s navigation pipeline.

Jun 2026 Education

Completed NPFL139 — Finished the graduate Deep Reinforcement Learning course at the Faculty of Mathematics and Physics, Charles University.

Sep 2025 Award

2nd Place at Robotour 2025 — Our team placed 2nd in the international outdoor robot competition with an autonomous beer-delivery robot.

Jun 2025 Research

Started PhD at CTU VRAS Group — Joined the Vision for Robotics and Autonomous Systems group, focusing on physics-guided learning for terrain-aware robot navigation.

Jan 2025 Publication

Master's thesis published — Unsupervised Approaches for 4D Panoptic Segmentation in LiDAR Driving Scenes. Introduces FlowSeg4D, achieving LSTQ 46.9 on SemanticKITTI without supervision. Read more →

Robotic Systems

Control algorithms for autonomous robot navigation based on predicted terrain properties

Autonomous Driving

Real-world LiDAR processing for robust object detection and tracking across diverse driving environments

Computer Vision

Multi-sensor fusion combining LiDAR, cameras, and depth sensors for robust perception

My Journey

BSc Cybernetics & Robotics

2020 - 2023

Czech Technical University
Thesis: Autonomous Road Crossing with a Mobile Robot

Porsche Engineering Student Contest

2022

2nd Place
Developed software for a 1:10-scale autonomous vehicle

MSc Cybernetics & Robotics

2023 - 2025

Czech Technical University
Thesis: Unsupervised Approaches for 4D Panoptic Segmentation in LiDAR Driving Scenes

Robotour

2025

2nd Place
Autonomous robotic beer delivery challenge

Researcher and PhD student

2025 - Present

Czech Technical University
Developing terrain-aware control and perception systems for autonomous robots

Robotour

2026

1st Place
1st place in the main category; lead integrator of the navigation pipeline

Technical Expertise

Core Languages

Python C++ LaTeX ROS2 Bash MATLAB

Computer Vision

PyTorch Open3D OpenCV

Tools & Platforms

Linux Vim Git Docker Gazebo NVIDIA Jetson

Current Research

Dissertation

Physics-Guided Learning for Machine Control
Control of autonomous robots based on terrain properties obtained through perception and physics-informed neural networks

PyTorch ROS2 Open3D Docker

Research Project

Outdoor Autonomous Robotics
Research on autonomous operations in unstructured outdoor settings, focusing on semantic segmentation, adaptive navigation, and localization

C++ Python NVIDIA Jetson HPC

Open Source

Online Robot Telemetry
Real-time telemetry and supervisory software for field-deployed autonomous robots

Open Science & Code
Most software packages, dataset tools, and benchmarks I develop are released under open-source licenses

Python JavaScript CI/CD Documentation

Personal Interests

Sports

Competitive ballroom dancer • Passionate runner, climber, and tennis player

Photography

Amateur photographer & videographer

Travel

Passionate traveler • Explored 15+ countries for academic conferences and leisure

Open to collaboration

I'm always interested in collaborating on autonomous systems research, whether it's joint publications, open-source projects, or industry applications.